Related Experiment Video
Updated: Apr 20, 2026
14:27
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
16.5K
Machine learning fMRI classifier delineates subgroups of schizophrenia patients
Maya Bleich-Cohen1, Shahar Jamshy2, Haggai Sharon3
1Functional Brain Center, Wohl Institute for Advanced Imaging, Tel Aviv Sourasky Medical Center, Tel Aviv, Israel.
Schizophrenia Research
|December 4, 2014
Summary
A novel machine learning method accurately distinguished schizophrenia patients with and without Obsessive-Compulsive Disorder (OCD) using functional MRI (fMRI) brain scans. This approach identified specific brain activation patterns linked to symptom severity.
Area of Science:
- Neuroimaging
- Psychiatry
- Machine Learning
Background:
- Identifying reliable neuroimaging markers in psychiatry is challenging.
- Schizophrenia presents complex clinical and methodological difficulties.
- Distinguishing schizophrenia subgroups requires advanced analytical methods.
Purpose of the Study:
- To apply a novel data-driven machine learning approach to functional MRI (fMRI) data.
- To delineate neural mechanisms differentiating schizophrenia patients with and without Obsessive-Compulsive Disorder (OCD).
- To identify robust fMRI-based classifiers for psychiatric subgroups.
Main Methods:
- Utilized functional Magnetic Resonance Imaging (fMRI) during a working memory task.
- Employed a data-driven multi-voxel pattern analysis (MVPA) approach called Searchlight Based Feature Extraction (SBFE).
- Analyzed data from 16 schizophrenia patients with OCD, 17 pure schizophrenia patients, and 20 healthy controls.
Main Results:
- The SBFE approach achieved 91% accuracy in classifying schizophrenia subgroups.
- Classification was based on activation patterns in the right intraparietal sulcus (r-IPS).
- r-IPS activation correlated with reduced symptom severity in schizo-obsessive patients.
Conclusions:
- The novel SBFE approach effectively differentiates symptom dimensions in complex psychiatric conditions.
- fMRI combined with machine learning offers a promising tool for psychiatric research.
- This method can aid in understanding neural mechanisms underlying comorbid psychiatric disorders.
